i-bond with Truflation

System_Architecture
bond dashboard screen

Inspiration

i-Bond (inflation linked bond) is very popular among retail working class people to protect their asset from inflation. However, currently issuance amount is not enough to meet people's demand. We would like to offer cost-efficient and transparent system for bond issuance and interest distribution. Then hopefully more i-bond will be allocated to more people by reallocating saved cost to interest rewards budget. In addition, inflation rate metrics should be more tightly tied to people's everyday life cost rather than metrics the central government defines, which sometimes affected by political factor.

What it does

Our system can offer following features to achieve quick, cost-efficient and transparent operations.

Create bond very easily by just specifying Currency, Maturity Period, Issued Amount.
Interest Rate is dynamically changed everyday and this will be the inflation rate defined by Truflation.
Accrued interest will be given to user every seconds and user could withdraw it at any time.
User can transfer their subscribed amount of bond to other user at their will instantly.

For inflation rate, we will take the metric calculated by Truflation in decentralized way. For the detail why central bank metric sometimes has problem and how Truflation can fix this problem is explained in their site. https://whitepaper.truflation.com/background/problem

How we built it

Please find the system architecture attached on image gallery. There are two highlight we can address on our system development.

1) Chainlink Integration
We have built custom job in order to fetch two parameters(inflation rate & Date) from truflation API. (Example job provided by Trulation retuned inflation rate only.) RateFetcher contract, which inherits ChainLinkClient, requests chainlink oracle to return "inflation rate" and "date", then verify if returned data is already fetched one or not by "date" value. If the data is new, RateFetcher contract will notify it to iBond contract (through iBondIssuer Contract). Lastly, rate update must be done every day. (Truflation rate is updated daily). We will reply on the ChianLink Keeper mechanism for this.

2) IBond Implementation
Core logic of i-bond is implemented in IBond contract, which is a bit complicated. IBond contract instance will be created whenever issuer create different i-bond(maturity period, currency, supply amount). IBond contract store all basic bond information as well as subscription situation of each subscriber. All necessary functions for subscriber (subscribe, claim accrued interest, transfer) are also supported in this contract.

What we learned

1) Chainlink technical framework
Learning Chainlink product is one of our objective to participate to this hackathon. Although we reply on node operator to host our external adapter on our hackathon product, I have researched the specification during hackathon and launch local chainlink node on my machine and host simple get->uint256 adapter by defining toml file. This work will give me fundamental knowledge to understand what chainlink can do and how it should be used.

2) Staking Model for i-bond implementation
When we implement core logic of i-bond, we learned the industry standard staking model, which is originally launched by Synthetix team, then customized it for our use case. Hackathon seminar "Defi Staking App: Code Along" was also very helpful for us to understand how this model works and how it can be tested.

What's next for i-bond with Truflation

1) Limit subscription amount per person
Usually, amount of i-bond each person can buy is limited due to its popularity and we should add function of it. For example, bond issuer can set "cap" amount per person, then they can change at any time depending on the subscription situation. (i.e. if less people buy, they can increase the cap.) Note if we seriously implement, perhaps, we should tied up with some crypto KYC vendor, as currently it is very normal for one person to have multiple ETH accounts. our system only can differentiate the account at the moment.

2) UI
Minimal UI is developed already. However it can be better. In addition, some test tool UI is nice-to-have. For example, reward distribution tool etc.

3) Code Improvement
A couple of TODO notes are added inside source code where we have room to improve

i-bond is, in nature, non-profitable financial product for the issuer. (To the contrary, it is very popular among subscribers.) So, the issuer is usually government, but I am aware that government will be cautious to be early adapter of new technologies...I wish some organization kindly shows an interest of PoC launch. Then we could have chances to brush up the product more.

Acknowledgments

For working on ChainLink integration with Truflation, I got a big help from professional people of these products. I would like to special thank for following people.

@joequant: Joseph at Truflation who explained the technical specification of Truflation API.
@mgladson: Matt at Block Farms who supported external adapter development of mine.